Provider-facing decision support for addiction treatment

Disengagement doesn't wait for discharge.
Neither can we.

Provider-facing decision support for addiction treatment. Divinia Health helps care teams recognize observable changes in engagement and follow through, without replacing clinical judgment.

The gap

Participation can begin to change long before a crisis or discharge. Yet the relevant information is often fragmented across attendance records, staff observations, and handoffs, making it difficult for care teams to recognize a pattern and respond consistently.


How it works

Three moves, in order.

01

Capture observable participation

Bring attendance and engagement information into one usable view.

02

Apply transparent program rules

Surface patterns based on criteria the provider can inspect, configure, and override.

03

Route and document follow-through

Notify the appropriate staff member and preserve an accountable record of the response.


What it looks like

One flag, followed all the way through.

Illustrative workflow using synthetic data.

Observable activity

Observe

Attendance, participation, and engagement events appear as structured data points.

Patient AOn track
Patient BFlagged
Patient COn track
One continuous view

Organize

Separate signals converge into one coherent operational picture.

Change requires attention

Surface

A meaningful change becomes visible, without being labeled as a prediction or diagnosis.

Provider-defined attendance rule triggered
Right person, clear next step

Route

The concern moves to a clearly identified staff owner.

Care coordinator notified.
Action: outreach call scheduled.

Action documented

Document

The action and follow-through are recorded, closing the operational loop.

10:42 AM: Call completed.
Follow-up documented.

Principles

Built around clinical judgment.

Clinicians stay in the loop

Divinia Health informs care. It doesn't automate it away. No flag should drive a high-stakes decision on its own.

Designed around provider data control

Built for the clinics and programs actually providing care, not around them.

Transparent by design

Every signal traces back to something a clinician can see, question, and override.

Privacy-first, by default

De-identified and synthetic data come first. Identifiable patient data is used only under appropriate governance, security controls, and data agreements.


Right now

We're developing and testing this workflow using synthetic and de-identified data, in preparation for a pilot inside a live behavioral health treatment program, under appropriate clinical governance, security controls, and data agreements.

Get in touch

Let's talk.